1. Technical Field
The present disclosure refers to a control device for a power factor correction device in forced switching power supplies.
2. Description of the Related Art
The use of devices for active power factor correction (PFC) is generally known for forced switching power supplies used in the electronic appliances of common use such as computers, televisions, monitors, etc. and for the supply of fluorescent lamps, that is of forced switching pre-regulator stages whose task is to absorb from the line an almost sinusoidal current in phase with the line voltage. Therefore a forced switching power supply of the current type comprises a PFC and a direct current to direct current converter or DC-DC converter connected to the output of the PFC.
A forced switching power supply of the traditional type comprises a DC-DC converter and an input stage connected to the electricity distribution line made up of a full wave diode rectifier bridge and by a capacitor connected immediately downstream so as to produce a non-regulated direct current starting from the alternating sinusoidal line voltage. The capacitor's capacitance is large enough to ensure that at its terminals a relatively small ripple is present in relation to a direct level. The rectifier diodes of the bridge, therefore, will conduct only for a small portion of each half-cycle of the line voltage, given that the instantaneous value of this is lower than the voltage on the capacitor for the greatest part of the cycle. The consequence is that the current absorbed from the line will be formed by a series of narrow pulses whose width is 5-10 times the average resulting value.
This presents considerable consequences: the current absorbed by the line has much greater peak and root-mean-square (RMS) values in comparison to the case of absorption of sinusoidal current, the line voltage is distorted by effect of the nearly simultaneous impulsive absorption of all the utilities connected to the line, in the case of three-phase systems the current in the neutral conductor results much increased and there is a low utilization of the energetic potential of the electricity production system. In fact, the waveform of impulsive current is very rich with uneven harmonics which, even though not contributing to the power given to the load, contribute to increasing the effective current absorbed from the line and thus to increasing the dissipation of energy.
In quantitative terms all this can be expressed both in terms of power factor (PF), intended as ratio between the real power (that which the power supply gives to the load plus that dissipated internally in the form of heat) and the apparent power (the product of the effective line voltage by the effective current absorbed), and in terms of total harmonic distortion (THD), generally intended as percentage ratio between the energy associated with all the higher order harmonics and that associated with the fundamental harmonic. Typically, a power supply with capacitive filter has a PF between 0.4-0.6 and a THD exceeding 100%.
A PFC, placed between the rectifier bridge and the input of the DC-DC converter, permits the absorption from the line of a nearly sinusoidal current in phase with the voltage, making the PF near 1 and reducing the THD.
In FIG. 1 a pre-regulator stage PFC is schematically shown comprising a boost converter 20 and a control device 1, in this case the control device L6563 produced by STMicroelectronics S.p.A. The boost converter 20 comprises a full wave diode rectifier bridge 2 having in input an alternating line voltage Vin, a capacitor Cin (that serves as filter for the high frequency) having the terminals connected to the terminals of the diode bridge 2, an inductance L connected to a terminal of the capacitor Cin, a power MOS transistor M having the drain terminal connected to a terminal of the inductance L downstream of the latter and having the source terminal coupled to ground by means of a resistance Rs suitable for enabling the reading of the current that flows in the transistor M, a diode D having the anode connected to the common terminal of the inductance L and of the transistor M and the cathode connected to a capacitor Co having the other terminal connected to ground. The boost converter 20 generates in output a direct current Vout on the capacitor Co, which is the input voltage of a user stage connected in cascade, for example a DC-DC converter.
The control device 1 maintains the output voltage Vout at a constant value by means of a feedback control action. The control device 1 comprises an operational error amplifier 3 suitable for comparing a part of the output voltage Vout, that is the voltage Vr given by Vr=R2*Vout/(R2+R1) (where the resistances R1 and R2 are connected in series with each other and in parallel to the capacitor Co) with a reference voltage Vref, for example of the value of 2.5V, and suitable for generating an error signal Se proportional to their difference. The output voltage Vout presents a ripple at a frequency that is double that of the line and superimposed to the continuous value. If however the bandwidth of the error amplifier is considerably reduced (typically lower than 20 Hz) by means of the use of a suitable compensation line comprising at least one capacitor and assuming an almost stationary regular operation, that is with constant effective input voltage and output load, this ripple will be greatly mitigated and the error signal will become constant.
The error signal Se is sent to a multiplier 4 where it is multiplied by a signal Vi given by a part of the line voltage rectified by the diode bridge 2.
At the output of the multiplier 4 a signal Imolt is present given by a rectified sinusoid whose width depends on the effective line voltage and on the error signal Se. Said signal Imolt represents the sinusoidal reference for the modulation PWM. Said signal is placed in input to the non-inverting terminal of a comparator 6 at whose inverting input the voltage present on the resistance Rs is proportional to the current IL.
If the signals in input to the comparator 6 are equal the same comparator 6 sends a signal to a control block 10 suitable for driving the transistor M and which, in this case, causes its turning off; therefore the output of the multiplier produces the peak current of the MOS transistor M which is enveloped by a rectified sinusoid.
After the transistor M has been turned off the inductor L discharges the energy stored in it on the load until it is completely emptied. At this point, the diode D opens and the drain node of the transistor M remains floating, therefore its voltage tends to the instantaneous input voltage through the resonance oscillations between the stray capacitance of the node and the inductance of the inductor L. Thus we see a rapid diminution of the voltage on the drain terminal of the transistor M that is sent in input to a device for detecting the passage through zero 13 through the auxiliary winding of the inductor L. The device 13 commands the turning on again of the transistor M, thus starting a new switching cycle.
The current absorbed from the line will be the low frequency component of the current of the inductor L, that is the average current per switching cycle (the switching frequency component is almost totally eliminated by the line filter placed at the input of the boost converter stage, always present for the electromagnetic compatibility regulations). For evident geometric reasons, the average current of the inductor is equal to half of the envelope of the peaks, and thus has a sinusoidal trend.
The multiplier 4 adjusts, by means of the error signal, the value of the sinusoidal reference for the PWM modulation upon variation of the load conditions and of the line voltage. In particular, considering the variations of the effective line voltage, if it, for example, doubles, the peak value also doubles; if the load does not change, and thus the power absorbed is constant, the input current, both the effective and the peak, once the transitory phase is over, halves in relation to the value that it had previously. The sinusoidal reference, nevertheless, is taken right from the rectified line voltage that is doubled. If the error signal did not intervene to correct the reference of the current (that is, if the regulation loop was open and thus the error signal was manually fixed), this would also become double (instead of half), thus giving place to a transfer of power four times greater. As the power requested by the load is constant, it would result in a considerable increase of the output voltage. The control loop, instead, reacting to this tendency, diminishes the value of the error signal so that the output of the multiplier becomes half of what it was previously.
Therefore the gain of the power block of a pre-regulator PFC depends in a quadratic manner on the line voltage and the error amplifier intervenes heavily to set the sinusoidal reference for the PWM modulation at the correct value independently from the line voltage.
Apart from the difficulties of planning the error amplifier, this strong dependence of its output voltage on the input voltage of the pre-regulator presents considerable consequences on the system. In first place, the quadratic variation of the gain of the power part implies a similar variation of the cutoff frequency of the open loop transfer function. If, then, the error amplifier is compensated to have 20 Hz band for the open loop transfer function at maximum line voltage, the band will be about 2 Hz at minimum line voltage, with the result of having an even slower dynamic response. In second place, by effect of the narrow band, the transient responses to sudden variations of the line voltage and of the output load will be very poor and there can be peaks of high voltage, limited only by the output dynamics of the multiplier, which is of the sinusoidal reference. These dynamics are set in such a manner that the maximum power requested by the load can pass to minimum line voltage, but this means that at maximum line voltage the pre-regulator is capable of carrying a power at least three times greater.
Finally, the fact that the output voltage of the error amplifier diminishes at the increase of the line voltage has a negative impact of the input current on the THD. In fact it can be demonstrated that the distortion of third harmonic introduced by the residual ripple superimposed at the continuous value present at the output of the error amplifier (whose gain at 100 Hz, for as much as it is low, is null) is proportional to the ratio between the peak-peak width of said ripple and the continuous value. The peak-peak width of the ripple is constant upon the variation of the line voltage, while the continuous value diminishes, thus the distortion of third harmonic increases.
These problems are usually solved by introducing in the control loop a feedforward of the line voltage and an inverter-squarer block (1/V2) like that included in the marked box of FIG. 1. In input to the multiplier 4 there is therefore a signal in output from an inverter-squarer block 41 at whose input is present a voltage signal Vff representative of the root-mean-square value of the line voltage obtained by means of a block 42; the signal in output from the block 41 is 1/Vff2. The function of this circuitry is that, in the first place, to generate a continuous level of voltage representative of the effective line voltage and, in second place, to use said level to adapt the output voltage of the multiplier to the variations of the input voltage without moving the output of the error amplifier.
This voltage representative of the effective line voltage is generated by means of a circuit detecting the peak of the voltage V1 that comprises a diode and a capacitor Cff.
To eliminate the detection error caused by the direct fall of the diode use is made of a so-called “ideal diode”, provided by interposing an operational amplifier connected to a non-inverting buffer and including the diode in the feedback. The capacitor Cff is equipped with a discharging means, that is, the resistance in parallel Rff so that the voltage at its terminals can adapt itself to the diminishing of the effective input voltage. This discharge, however, should be imperceptible in the environment of each half line cycle, so that the voltage at its terminals is, as much as possible, close to continuous. With the above mentioned conditions and considering the capacitance and resistance values that can be obtained in integrated form, it is convenient for the Rff and Cff to be elements placed outside the integrated control circuit.
However in the case of sudden drop in the line voltage, the system in FIG. 1 replies with an exponential trend having a time constant Rff*Cff which, for what was said, will be to the order of many hundreds of milliseconds. This leads to the feedforward system losing effectiveness for a time which is as long as the variation of the input voltage is large and is as long as the time constant Rff*Cff. In fact, even though the signal on the comparator 6 tends to increase, the signal Vff is still too high because of the slow discharging and the output of the multiplier cannot adapt itself to the new level of current in input requested. The result is that the error amplifier tends to go out of its range and its output to saturate high. This causes a deep undershoot in the output voltage that can carry the converter downstream, fed by the stage PFC, out of regulation.